Well, I can only tell you how I’m doing it.
For first drafts, I try to tie my hands away from anything but the keyboard, and do my best (and fail bitterly) NOT to look things up on the ‘net. However, once the raw material is down, then I can insert things like specific bird and tree names, or expand on a situation or scene. I did very little research for most of the stories in my short story collection A Sort Of Walking Miracle. I did very little research for first draft of my novella, Serendip (draft title). Now, as I’m developing it to novel length, I’m finding I am doing some research as I go along. I’m reading about the period the work is set in, both in fiction and non fiction. Doing this has actually inspired further scenes in the book and because of this I have had to alter the original time frame.
I will be going to Sri Lanka for a two week trip at the end of August for hands on research. I hope to interview people who experienced some of the events that occur in my novel, and I’ll be taking in the surroundings and environment as much as I can. The last time I was in Sri Lanka was in the early 80s, and there is much I have forgotten. I thought I hated research, but now I’m actually enjoying it.
I have read of writers who do a lot of research before they sit down and write. Perhaps this will be an approach for my next novel, right now I feel I’m very much learning the ropes as I go along. Perhaps it depends on the story. I do know that I once destroyed a good story idea by getting lost in Google–I wrote 40,000 words and had to throw the towel in because I had written far too many ‘interesting’ words and had zero plot or character development.